PEDS 2007

Analysis and Comparative Study on the Performance between Standard and High Efficiency Induction Machines operating as Self-Excited Induction Generators
B. Sawetsakulanond and V. Kinnaraes
Dept. of Electrical Engineering, King Mongkut's Institute of Technology Ladkrabang, Bangkok 10520, Thailand Fax 662-3269902 E-mail:Budhapon dghotmail.com, E-mail: kkwijit dWkmitl.ac.th

Abstract--This paper proposes the performance analysis and comparative study between standard and high efficiency squirrel cage self - excited induction generators (SEIG). Consideration of capacitance of SEIG based on equivalent circuit model is given. Testing and performance comparison under dynamic and steady state conditions for both pure resistive and resistive - inductive load conditions have been made. The interpretation of experimental results can be guidelines for effective design and development of SEIG for wind energy applications.

analysis of capacitance with the proposed approach depending on load conditions and comparative study on the perfornance between standard and high efficiency machines operating as self- excited induction generators. The results will be graphically presented and discussed in order to provide guidelines for the designers.
II. EQUIVALENT CIRCUIT ANALYSIS
For proposed capacitance consideration of the SEIG, the system can be shown in Fig 1. Capacitors are divided into two parts namely, a built-up or excitation capacitor (Cb) and a compensating capacitor (Cc) for terninal voltage regulation. The Cbis responsible for no-load operation whilst both Cb and Cc are responsible for onload operation. For no-load operation, the per phase equivalent circuit is shown in Fig2. neglecting harmonic effect and core loss[3].
- - - - -

Index Terms--self - excited induction generator, induction machines

I. INTRODUCTION Owing to the continuous increase in energy need, it is difficult to meet the growing demand by exploiting energy from the limit conventional source, such as coal, oil, gas, and so on. As a consequence, a greater emphasis is now being given to harness energy from non conventional sources such as wind , biogas and small hydro heads[1,3]. A three phase induction machine can be made to work as a self - excited generator (SEIG) when its rotor is driven at suitable speed by wind energy and its excitation is provided by connecting a three phase capacitor bank at the stator terninals in order to build-up voltage and regulate terninal voltage. It offers various advantages over other machines such as reduced unit cost, brushless rotor (squirrel cage construction), absence of separate DC source and ease of maintenance. Numerous papers have attempted to analyze the SEIG using equivalent circuit approach [1-5]. Generally two different (but related) methods of capacitance value solution for voltage build-up and terninal voltage regulation have been employed, namely, the loop impedance method and nodal admittance method. Steady state analysis of such generators seems to be more interested than dynamic analysis. Excitation capacitors affects stator current. Therefore careful selection of this capacitor is required. Controlled static var compensator in conjunction with the ac load voltage regulator based on dynamic model can be found in [6]. These techniques are based on power electronic applications which provides good perfornance under a wide range of operation. However there are a few research works on investigation of various types of squirrel cage induction machines with suitable capacitance subject to load conditions based on separation of capacitors into built-up capacitor and compensating capacitor. Therefore this paper present the

- -

- -

Vt

IC1
Induction Gen 14

I 4cc -c

Load

No- Load 1 JjC On-- LoadIl I._._._._._._._._.j 1._-_. --_-_-_-_-_-_-_I

Fig. 1. Single line diagram of the SEIG

From Fig.2, impedance, ZCD can be written as equations

(1)- (3)
ZCD RCD + jXCD
RCD CD

(1)
(2)

R22X+ (a b)2 (X + X2()2

-

(a b)R2Xm2

XCD -

R22X

R22 + (a b)2 (X. + X2)2

+ (a-b)2 X

X2(X. + X2)

(3)

Where a is the per unit frequency b is the per unit speed

1-4244-0645-5/07/$20.002007 IEEE

1313

For analysis of the on - load operation, Cc is introduced. According to Fig.3, when supplying the resistive - inductive load, the terminal voltage level will be reduced since the load draws reactive power from the system. In order to maintain the terminal voltage constant, Cc can be determined as follows. Total reactive power of overall capacitors (i.e. Cb and Cc) is
ZAB

Fig.2. Per phase equivalent circuit of the SEIG at no - load

--r----------Qg

r--------------VtQ

Using KVL results in

I

Induction GenQcbt

tQcc

Load

Jcb 2

+JiXl )+ZCDj

(4)

iC1
Load

il

pC

On-Load

Since Ig is not definitely zero during voltage buildup, as a consequence, equation (4) can be rewritten as

Fig.3. Power flow diagram

Q,t
=

+RI + JX1) + ZCD

(5)
When

Qg + QL

(10)

From equation (5), the real part is zero. Then

Qg

=CSg2

p2

(1 1)
(12)

RI +
a

(a - b)R2 Xm2 R22 + (a - b)2 (Xm + X2 )2

(6)

According to equation (6), we can obtain

QL=XL 2
amax as

amax

bb

(7)

Where SG is the per phase apparent power of the SEIG PG is the per phase active power of the SEIG QG is the per phase reactive power of the SEIG QL is the per phase reactive power of load Q, is the per phase reactive power of total capacitor

Also, bh can be determined as

b
Where

Compensated reactive power is

(8)
When
Qcc = Qct Qcb
Qcb
=

WXm RisXm

2RI R2 +(1+X2J

(13)

bc

is the critical speed

27Q() 2(fCb)

(14)

Therefore capacitance for built-up voltage during no-load can be determined as Where QCb iS the per phase reactive power of the built - up capacitor 1 Q, is the per phase reactive power of the = ( Cb b (9) compensating capacitor x ITfb Z b . max2 X1+X D) Where Cb is the per phase built - up capacitor value Zb is the base impedance fb iS the base frequency
Per phase current of the compensating capacitor (Cc ) is determined from
_ QCC Icc -

vtI Za

(15)

1314

Per phase compensated capacitance is calculated as following equations

la X cc = Icc

Table 2 Capacitance for resistive load

NoType

ON-Load; Cc (,iF)
192 384 576 768 Total

vt

Load

CT
(piF)
68
58

(16)
Standard

c,
(

(F) W/ph)
5
3

(W/ph)
7
4

(W/ph) 9
5

(W/ph) 12
6

Cc (,F)

35
40

33
18

cc

2rf (iX cc)

(17)

Hi-efficiency

Thus, total capacitor value for the SEIG is

Table 3 Capacitance for resistive - inductive load

No-

ON-Load; Cc (,iF)
213 424 648 859 Total

Ct Cb +Cc
=

(18)

Type

Load

CT
(piF)
110 99

(pF)

C,
35

(W/ph) 14 10

(W/ph)

(W/ph) 20 16

(W/ph) 24 20

Cc (,F)

Where

cc

is the per phase compensating capacitor Cb is the per phase built - up capacitor Ct is the per phase total capacitor

Standard

17
13

75
59

Hi-efficiency

40

The procedure for determining capacitor values for the SEIG under on- load conditions when supplying the resistive load can be performed as same as for the resistive - inductive load with QL =
III. EXPERIMENTAL TESTS AND DISCUSSION

The tests have been made for finding parameters and for investigation performance
A. Parameters Test Three phase, 3kVA, 2.2 kW, 220/380 volt, 4 poles Fig.4. Characteristic between Eg and Xm induction machines with standard and high efficiency types are used. Table 1 shows parameters of the under It is true that in the design of the high efficiency machine, test machines obtained from the test complied with IEEE maximum flux density level is low in order to reduce core std. 112-1996 testing [7]. loss. As a result, Xm is low. For SEIG applications of the high efficiency induction machine, Cb for the high Table 1 efficiency SEIG is higher than for the standard SEIG. Parameters of the Machines
Type
RI
3.17
R2
Rc

Xi
3.48

X2

Xm

2.56

627.37

3.48

93.43

Standard Machine
High Efficiency

QQ Q
1.40

Q
1.81

Q
568.50 3.31

756 331

Machine

B. Operating Test The capacitance analysis uses Maple program for determining capacitance under various load conditions and 0.85 lagging power factor. Tables 2-3 give results with constant speed of 1500 rpm and regulated terminal voltage of 220 V, Y connected. Figs. 4-16 show the SEIG performance in terms of voltage build - up, power quality, frequency variation and efficiency. Fig.5. Variation of speed with terminal voltage for SEIG From Fig.4, it can be seen that characteristics between Eg and Xm for both SEIGs are different. Xm of the Figs.5-6 demonstrate the impact of the speed change on high efficiency SEIG is smaller than that of the standard built- up voltage. When increasing the speed of the prime mover, the high efficiency SEIG has a critical speed SEIG.

1315

point, which is lower than the standard SEIG. When decreasing the speed of the prime mover, built- up voltage
of of the

high efficiency SEIG is longer than since the high efficiency SEIG magnetic
flux

for the standard SEIG is

designed

for

low

high efficiency

SEIG is decreased lower

density.

reason is that in the design high efficiency machines, magnetic material type with narrow hysteresis loop is always selected in order to reduce the hysteresis losses. As a consequence, the

than the standard SEIG. The

~MX4LNM

shis/ay

capability
view

of

voltage regulation
the
to

is influenced with this

SEIG

point, improvement

high

overcome

this

efficiency problem.

requires

aTraWla

P-P F rom

rw. BY So ijmz

lm*

RUK

22

Fig.9.
under

Terminal

voltage
,Oo

waveform
THDv

of the standard SEIG

steady

-state

=4.7
5ms/div
m

CH1=200V DC 100:1

~~~~~~NORM:200kS/s

~~TracB1632.OV P-P

Max.......

...............

Fig.6.

Variation of

hysteresis voltage

for SEIG

Fr.....q..50...0.....z

Fig.

Terminal

voltage

waveform
,

of the
6

high efficiency

SEIG under

steady-

state

THDv=

250-

200

150

Fig.7.

Built -up

voltage

of the standard SEIG

50-

100 1-1

E3 ~~~~~~~Standard

15

21

Harmonic order

Fig. Figs.
than

1.

Harmonic

Spectra

of the SEIG terminal

voltage

9-1 1

show

terminal
more

standard SEIG offers the

voltage waveforms. The nearly sinusoidal waveform

high efficiency SEIG. Fig. 1. confirms this point. According to this disadvantage the high efficiency SEIG should be improved. Figs. 12-13 show variation of
reactive power with

Fig.8.

Built- up

voltage

of the

high- efficiency

output power under various load

standard and of
a

SEIG

conditions.

Q,b

of the

high efficiency
SEIG

is

decreased due to the

impact

reduction in

frequency
is

Figs.7-8
of

illustrate to duration

(built-

up

voltage time, tb)

of induced

voltage.

Q,b of the

high efficiency

voltage

build- up of the SEIG. The duration for the

higher

than that of the standard SEIG due to lower

X,,m*

1316

As a result, it needs higher capacitance. QCc is increased with increasing loads. Qccof the high efficiency SEIG under pure resistive load and resistive - inductive loads. As a consequence, it needs lower compensated capacitance since the high efficiency SEIG has voltage drop due to low stator impedance and low QG. The sum of reactive power of total capacitors is equal to the sum of reactive power of the SEIG and loads.
law.
a

Ifi

-M

._

ri.

;G'F

Rui~~~~~~~~~~~~~~~~1i~~~~~~~~~~uLo~~~~~~~~~~~~~~~~~~~td~~~~~

Fig. 15. Variation of frequency with output power for resistive inductive loads
-

.~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ . . . . .
Qahll

.#q

Z"

S"

440

"

70

power

Fig. 12. Variation of reactive power with output for resistive load

vt 4o

Fig. 16. Variation of efficiency with output power for both SEIG
From Fig. 16, it can be seen that the efficiency of the high efficiency SEIG is higher than that of the standard SEIG under various types of loads. The fact is that in the design of the high efficiency machine, a reduction in losses is the main objective. As a result, the high efficiency machine still produce low losses when operating as the SEIG.

Fig. 13. Variation of reactive power with output power for resistive inductive loads
-

Figs.14-15 show a change in frequency of the SEIG. The high efficiency SEIG gives smaller change than the IV. CONCLUSION standard SEIG under various types of loads since the rotor resistance of the high efficiency SEIG is less than From the tests and analysis of the performance of that of the standard SEIG. the high efficiency and standard SEIG, it found that the high efficiency SEIG is suitable for generating electric power from wind energy. However, improvement and development are still required in term of capability of voltage regulation under a speed change, harmonic reduction of terminal voltage and start - up duration
APPENDIX
Vb: 220 V, lb: 5.2 A (standard machine) and 4.7 (high efficiency machine), Pb: 2.2 kW, Zb: 42.30 Q (standard machine) and 46.80 Q (high efficiency machine) ZL: per phase resistive load 242 Q and resistive- inductive load 166.97+j 103.45 Q
,
.

4_

Fig.

14.

Variation of

frequency

with output power for

resistive load

1317

REFFERENCES
[1] S.S. Murthy, O.P. Malik, and A.K. Tandon, " Analysis of self- excited induction generators," Proceedings of the IEE, Vol.129, No.6, pp.260-265, November 1982. [2] S.S. Murthy , B.P. Singh , C. Nagmani and K.V.V. Satyanarayanna," Study on the use of conventional induction motors as self excited induction generators," IEEE Transactions on Energy Conversion, Vol.3, No.4, pp.842-848, December 1988. [3] T.F. Chan , "Capacitance Requirements of Self-exited IEEE Transactions on Energy Induction Generators," Conversion, Vol.8, No.2, pp.304-3 10, June 1993. [4] S.P Singh, B.S Singh and M.P. Jain, "Comparative study on the performance of a commercially designed induction generator with induction motors operating as self excited induction generators," Proceedings ofthe IEE, Vol.140, No.5, pp.374-380, September 1993. [5] L. Shridhar, Bhim Singh, C.S. Jha , "A step towards improvements in the characteristics of Self excited Induction Generator ," IEEE Transactions on Energy Conversion, Vol.8, No. 1, pp.40-46, March 1993.

Polyphase Induction Motors and Generators ," pp.28-53,

[6] Tarek Ahmed, Katsumi Nisshida, Mutsou Nakaoka and Hyun Woo Lee, "Self - Excited Induction generator with Simple Voltage Regulation Scheme for Wind Energy," Proceedings of the 30 th Annual Conference of the IEEE, Vol.140, No.5, pp.86-91, November 2004. [7] IEEE std 112 - 1996, "IEEE Standard Test Procedure for
1996.

1318